Pipeline for retrieval of COVID-19 immune signatures

Objective: The accelerating pace of biomedical publication has made retrieving papers and extracting specific comprehensive scientific information a key challenge. A timely example of such a challenge is to retrieve the subset of papers that report on immune signatures (coherent sets of biomarkers) to understand the immune response mechanisms which drive differential SARS-CoV-2 infection outcomes. A systematic and scalable approach is needed to identify and extract COVID-19 immune signatures in a structured and machine-readable format. Materials and Methods: We used SPECTER embeddings with SVM classifiers to automatically identify papers containing immune signatures. A generic web platform was used to manually screen papers and allow anonymous submission. Results: We demonstrate a classifier that retrieves papers with human COVID-19 immune signatures with a positive predictive value of 86%. Semi-automated queries to the corresponding authors of these publications requesting signature information achieved a 31% response rate. This demonstrates the efficacy of using a SVM classifier with document embeddings of the abstract and title, to retrieve papers with scientifically salient information, even when that information is rarely present in the abstract. Additionally, classification based on the embeddings identified the type of immune signature (e.g., gene expression vs. other types of profiling) with a positive predictive value of 74%. Conclusions: Coupling a classifier based on document embeddings with direct author engagement offers a promising pathway to build a semi-structured representation of scientifically relevant information. Through this approach, partially automated literature mining can help rapidly create semi-structured knowledge repositories for automatic analysis of emerging health threats.

Disparity in biomedical publication trends in the CARICOM countries: Is there a need for a more vibrant research culture in the region?

Background: Biomedical research and publications provide evidence-based information about the extent and burden of health-related problems of a country and help to formulate strategic and operational plans to tackle the problems. Purpose: To determine the biomedical publication rates of CARICOM full member countries. Methods: Biomedical publications of full member CARICOM countries were retrieved using PubMed (1990–2015) and SCImago Journal & Country Rank (1996–2015) databases. CARICOM countries having >50 publications in the PubMed (1990–2015) database were subject to further analysis, whereby publications of each country were adjusted by total population (million population), gross domestic product (billion-dollar), and Internet usage rate (hundred thousand population). Results: Total publications by all countries were 7281 and 8378 in PubMed and SCImago Journal & Country Rank, respectively. Jamaica produced highest number of publications (PubMed: 3928 (53.9%); SCImago Journal & Country Rank: 2850 (34.0%)). In both databases, Grenada had the highest research publications when adjusted with per million population (4721 and 10,633), per billion gross domestic product (803 and 1651), and per hundred thousand Internet users (1487 and 3387). Trend analysis revealed Jamaica produced the highest number of additional PubMed listed publications each year, averaging 4.8/year, followed by Trinidad and Tobago (4.4). According to SCImago Journal & Country Rank, Jamaica also had the highest number of citations (42,311) and h-index (76), followed by Trinidad and Tobago (29,152 and 71). Barbados had the highest number of citations per document (24.9), followed by Haiti (18.4). The publication rates determined by PubMed and SCImago Journal & Country Rank databases were significantly correlated (p < 0.001). Most publications (68% SCImago Journal & Country Rank and 85% PubMed) can be attributed to authors affiliated with Barbados, Jamaica, and Trinidad. Conclusion: Publication and citation rates varied markedly between CARICOM countries and were in general low. Most publications could be attributed to researchers affiliated with The University of the West Indies. More universities valuing biomedical research are needed in the region, and more resources needed to improve publication rates.

Perspectives From Authors and Editors in the Biomedical Disciplines on Predatory Journals: Survey Study

Background Predatory journals fail to fulfill the tenets of biomedical publication: peer review, circulation, and access in perpetuity. Despite increasing attention in the lay and scientific press, no studies have directly assessed the perceptions of the authors or editors involved. Objective Our objective was to understand the motivation of authors in sending their work to potentially predatory journals. Moreover, we aimed to understand the perspective of journal editors at journals cited as potentially predatory. Methods Potential online predatory journals were randomly selected among 350 publishers and their 2204 biomedical journals. Author and editor email information was valid for 2227 total potential participants. A survey for authors and editors was created in an iterative fashion and distributed. Surveys assessed attitudes and knowledge about predatory publishing. Narrative comments were invited. Results A total of 249 complete survey responses were analyzed. A total of 40% of editors (17/43) surveyed were not aware that they were listed as an editor for the particular journal in question. A total of 21.8% of authors (45/206) confirmed a lack of peer review. Whereas 77% (33/43) of all surveyed editors were at least somewhat familiar with predatory journals, only 33.0% of authors (68/206) were somewhat familiar with them (P<.001). Only 26.2% of authors (54/206) were aware of Beall’s list of predatory journals versus 49% (21/43) of editors (P<.001). A total of 30.1% of authors (62/206) believed their publication was published in a predatory journal. After defining predatory publishing, 87.9% of authors (181/206) surveyed would not publish in the same journal in the future. Conclusions Authors publishing in suspected predatory journals are alarmingly uninformed in terms of predatory journal quality and practices. Editors’ increased familiarity with predatory publishing did little to prevent their unwitting listing as editors. Some suspected predatory journals did provide services akin to open access publication. Education, research mentorship, and a realignment of research incentives may decrease the impact of predatory publishing.

The Deep Learning–Based Recommender System “Pubmender” for Choosing a Biomedical Publication Venue: Development and Validation Study (Preprint)

BACKGROUND It is of great importance for researchers to publish research results in high-quality journals. However, it is often challenging to choose the most suitable publication venue, given the exponential growth of journals and conferences. Although recommender systems have achieved success in promoting movies, music, and products, very few studies have explored recommendation of publication venues, especially for biomedical research. No recommender system exists that can specifically recommend journals in PubMed, the largest collection of biomedical literature. OBJECTIVE We aimed to propose a publication recommender system, named Pubmender, to suggest suitable PubMed journals based on a paper’s abstract. METHODS In Pubmender, pretrained word2vec was first used to construct the start-up feature space. Subsequently, a deep convolutional neural network was constructed to achieve a high-level representation of abstracts, and a fully connected softmax model was adopted to recommend the best journals. RESULTS We collected 880,165 papers from 1130 journals in PubMed Central and extracted abstracts from these papers as an empirical dataset. We compared different recommendation models such as Cavnar-Trenkle on the Microsoft Academic Search (MAS) engine, a collaborative filtering–based recommender system for the digital library of the Association for Computing Machinery (ACM) and CiteSeer. We found the accuracy of our system for the top 10 recommendations to be 87.0%, 22.9%, and 196.0% higher than that of MAS, ACM, and CiteSeer, respectively. In addition, we compared our system with Journal Finder and Journal Suggester, which are tools of Elsevier and Springer, respectively, that help authors find suitable journals in their series. The results revealed that the accuracy of our system was 329% higher than that of Journal Finder and 406% higher than that of Journal Suggester for the top 10 recommendations. Our web service is freely available at https://www.keaml.cn:8081/. CONCLUSIONS Our deep learning–based recommender system can suggest an appropriate journal list to help biomedical scientists and clinicians choose suitable venues for their papers.

Automated Text Detection and Recognition in Annotated Biomedical Publication Images

Images in biomedical publications often convey important information related to an article's content. When referenced properly, these images aid in clinical decision support. Annotations such as text labels and symbols, as provided by medical experts, are used to highlight regions of interest within the images. These annotations, if extracted automatically, could be used in conjunction with either the image caption text or the image citations (mentions) in the articles to improve biomedical information retrieval. In the current study, automatic detection and recognition of text labels in biomedical publication images was investigated. This paper presents both image analysis and feature-based approaches to extract and recognize specific regions of interest (text labels) within images in biomedical publications. Experiments were performed on 6515 characters extracted from text labels present in 200 biomedical publication images. These images are part of the data set from ImageCLEF 2010. Automated character recognition experiments were conducted using geometry-, region-, exemplar-, and profile-based correlation features and Fourier descriptors extracted from the characters. Correct recognition as high as 92.67% was obtained with a support vector machine classifier, compared to a 75.90% correct recognition rate with a benchmark Optical Character Recognition technique.